[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Clarification concerning security of testing on a laptop



My 2 cents on this: I've been using unstable branches/testing for the
last couple of years now, and only once did I see broken packages that
stuck me there badly (you know, some combination of libc+perl+dpkg, or
something that lethal). Even then, with some fancy pussyfooting I
managed to repair everything without a fresh install.

One thing I still don't like about Debian is the time it takes to get
new versions of major packages, even on "unstable" or "testing" (think:
Gnome 2.8). The other side of that coin is that the package maintainers
DO keep their eyes on the target. Boy do they work efficiently !

Jeff

On Tue, 2005-01-04 at 19:26 -0500, Brian Pack wrote:
> On Tuesday 04 January 2005 06:26 pm, Kent West wrote:
> 
> > My personal opinion? Skip Testing and go straight to Sid. You have more
> > chance of breakage (although it's been very rare in my experience (about
> > 3 years now)), but said breakage also tends to get fixed within hours
> > instead of 10 days. Same for vulnerabilities. As soon as a vulnerability
> > is found, if the developer/maintainer of that package is on the ball,
> > the fix will be in Sid very quickly, perhaps even before it makes it
> > into Stable. You also get newer toys to play with.
> 
> I like toys. :)
> 
> I may not have the hands-on experience with Sid that you have (replaced SuSE 
> 9.1 with Sid in July... 99:1 linux:xp usage since), but I also have only seen 
> very rare breakages. I've even seen 'breakages' that weren't breakages at 
> all, but new packages that weren't as forgiving of existing errors as older 
> versions.
> 
> Case in point: The new Fluxbox 0.9.11. When I replaced my 0.9.9, the desktops 
> would now jump two at a time when I used the scroll wheel. I thought it was a 
> bug in fluxbox. Then I noticed that openbox was doing the same thing. 
> Judicious googling got me to a page that said to check my XF86Config. 
> Sometimes there are 2 separate listings for the mouse, and could be sending 
> double inputs. I edited out the redundancy, and bingo! No more double jumps.
> 
>  I also make it a point to subscribe to debian-security. The reports almost 
> always list fixes for both Woody and Sid.



Reply to: